According to Reuters, Intel’s autonomous driving company, Mobileye, has signed a contract to provide autopilot technology to a European automaker’s 8 million vehicles. However, the specific terms of the contract and the identity of the automaker were not disclosed.
Mobileye Test Vehicle (Source: Reuters)
Intel bought Mobileye for $15.3 billion last year, which is one of the largest transactions to date, signifying that automakers and suppliers are accelerating the landing of certain autonomous driving technologies. Although the current level of technology can only be applied to limited traffic Situations, such as freeway driving and emergency braking, will take several years to get from fully automated driving into the market, but this does not prevent manufacturers from earning profits from it.
Erez Dagan, Senior Vice President of Mobileye, stated that the contract for the application of advanced driver assistance systems will start in 2021. At that time, Intel's EyeQ5 chip designed for full autonomous driving will replace the EyeQ4 chip.
Of course, in such a large market, not only one company is involved. Mobileye is competing with several chip and machine vision system makers such as Nvidia for the opportunity to provide 'brains' and 'eyes' for self-driving cars.
Mobileye stated that of the cars currently on the road, about 27 million cars from 25 car manufacturers already use some type of assisted driving system, and Mobileye’s market share exceeds 70%.
Mobileye CEO Amnon Shashua stated, 'We expect that by the end of 2019, Mobileye will have installed more than 100,000 auto-piloted L3-class cars. 'L3-class vehicles can achieve autonomous driving, but when the system cannot continue to operate, the driver has 10 seconds to take over the car.
Mobileye is working with a number of automakers such as General Motors, Nissan, Audi, BMW, Honda, Fiat Chrysler and China Weilai Auto to provide them with L3 autopilot technology next year.
Technician Debugs Mobileye Auto Drive (Source: Reuters)
At the same time, at the Jerusalem headquarters, Mobileye also tested the more advanced L4 level autopilot technology on the Ford Fusion hybrid. The test vehicle was equipped with 12 small cameras and four upcoming EyeQ4 chips. The cars can be automatically driven on the freeway in Jerusalem noon, without driver intervention.
Mobileye stated that although the L4 system is expected to begin production in 2021, many of the technologies it contains are closely related to some consumer-grade systems that will soon be available.
Shashua said that car manufacturers have made a promise that driverless taxis (Robo taxis) should be able to go around 2021. 'When we designed the system, we considered not only all the technologies in front of us, but also the next year. , Technology within two years, then driverless taxi.
He pointed out that by then there will be a number of higher-priced luxury private cars, which may also include some mid-priced cars, using the same technology. However, each car needs an additional cost of about $12,000 for upgrading.
Shashua said that perhaps a few years later, the vehicles on the road will include ordinary cars and driverless cars. Because of this, safety is particularly important. 'Despite the fact that every year in the United States, 40,000 people are killed in traffic accidents, In driving a car, the public obviously will not tolerate so many casualties.
Therefore, Shashua said that self-driving cars cannot rely solely on cameras. To prevent accidents and make the system make the best driving decisions, the car needs to process combined data from cameras, high-definition maps, radar and laser radar.
'The test vehicle is based on the idea that a car can be driven like a human driver. In Jerusalem, the driving style is very decisive, because the driving culture here is very decisive.'
'You want to be safe enough and you want to be decisive,' Shashua added. If the driving style is too hesitant, it may make other drivers impatient and lead to accidents. 'In the future, the system will observe other driving on the road. , and adapt to the driving situation at that time, like a human driver.
In the case of designing a self-driving car, one question is how to define what is a dangerous situation. Our transport law is comprehensive, but it lacks a formal definition. 'We want to formalize these situations in advance so that the machine will not be put in a dangerous situation.' Shashua said. 'However, this issue may eventually require the court to resolve.'